3174
CH,
CH3
\
H
-N
,
GHs
I
1 CH3
/
,
(4) W. Theilacker and 0. R. Leichtle, Ann., 572, 124 (1951).
CONHCH,
C&
CONHCHJ I
V
(CDC13)N-CH3, 6 4.08; NHCH3, three-proton doublet centered at 6 3.05 (J = 5 cps); , , ,A (2-propanol) 222 (E 23,200), 260 (e 25,750), 298 ( E 5000), 306 (e 5050), and 356 mp (E 19,000). Anal. Calcd for C17HlsN20: C, 77.25; H, 6.10. Found: C, 77.59: H, 5.85. The isoindole structure of VI was proved by comparison with an authentic specimen, prepared from the known 2-methyl-1-phenylisoindole and methyl isocyanate. A detailed account of this investigation will be the subject of a forthcoming publication. Acknowledgment. We wish to thank Professor G. Buchi for many helpful suggestions and his valuable advice.
vI
223-228", mmp with IV 190-202'; Y ~ 3410 ~ cm-'; ~ ' CHCls Y ~ =1680 ~ and 1530 cm-I; nmr spectrum (CDCl,) NHCHa, three-proton doublet centered at 6 2.65 ( J = 5 cps), N H 6 6.55, 10 protons in aromatic region (band) centered at 6 7.5; , , A, (2-propanol) 253 mp (e 13,700). Anal. Calcd for CI6Hl4N20: C, 76.78; H, 5.64. Found: C,76.79; H,5.34. The isoindolenine I11 could be reconverted quantitatively to the thermodynamically more stable isoindole IV simply by treatment with dilute acid or strong aqueous base. The mechanism of the rearrangement can be explained by an initial attack of hydride ion at the 3 position of the benzodiazepinone nucleus (removal of one of the acidic protons) followed by a ring contraction t o give either of the tricyclic intermediates as shown. Either intermediate would then undergo further rearrangement to give the salt of the isoindole carboanion 11. The 1,4-diniethyl-2-oxo-l,4-benzodiazepiniumsalt V could also, by treatment with sodium hydride, be rearranged to the corresponding 2,N-dimethyl-3-phenyl1-isoindolecarboxamide (VI), mp 185-188'; &Zl33460 CHCla cm-I; vC=* 1640 and 1530 cm-I; nmr spectrum pound IV. These shifts indicate that the CONHCHBgroup is attached to a saturated center on the isoindolenine nucleus.
Journal of the American Chemical Society / 88:13 / July 5, 1966
R. Ian Fryer, J. V. Earley, L. H. Sternbach Chemical Research Department, Hoffinaiiri-La Roche Itic. Niitley 10, New J e n e y Receiced Muy 2 , 1966
The Pinacol Rearrangement of 2-p-Anisylnorbornane-2,3-cis,exo-dio11 Sir: Collins, et a1.2 have aptly demonstrated that the rearrangement of 2-phenylnorbornane-2,3-cis,exo-diol (I) in concentrated sulfuric acid at 0" takes place with rearrangement of the norbornane carbon skeleton cia a hydride shift from the 6 to the 1 p ~ s i t i o n ,ac~ companied by an intramolecular migration of hydrogen from C-3 to C-2 to give 3-endo-phenyl-2-norbornanone (11). Roberts4 has suggested that the 2,4-dimethoxy analog of I could give endo-3-hydrogen migration during the pinacol rearrangement, as it is conceivable that the two methoxy groups could stabilize the tertiary classical open carbonium ion I11 at the expense of the bridged ion IV. I n connection with another study in the norbornane system we had prepared 2-p-anisylnorbornane-2,3cis,exo-diol (V) and have carried out its rearrangement in concentrated sulfuric acid at 0". While some ~ sulfonation did take place, the only nonacidic compound isolable was 3-endo-p-anisyl-2-norbornanone (VI) in ca. 50% yield. The preparations of the compounds involved in this study and their structural assignments are as follows. Peracetic acid oxidation of 2-p-anisylnorbornene followed by lithium aluminum hydride reduction gave the diol V. The performic acid oxidation procedure which previously afforded I5 proved unsatisfactory as a pathway to V, apparently due to the more rapid alkene dimerization reaction.6 The structure of V was confirmed by comparison with I in the 3-p region of the infrared, where both showed strong intramolecular O H . 0 hydrogen bonds,' and
-
(1) This work was supported by a research grant from the Natlonal Science Foundation. (2) C. J. Collins, 2. I
